Object Tracking

  1. Fundamental Concepts and Components
    1. The Tracking Loop
      1. Prediction
        1. State Estimation Before Observation
          1. Use of Motion Models
          2. Update
            1. Incorporating New Observations
              1. Correction of State Estimates
            2. Object Representation
              1. Point Representation
                1. Centroid Tracking
                  1. Applications and Limitations
                  2. Bounding Box Representation
                    1. Axis-aligned Boxes
                      1. Rotated Bounding Boxes
                      2. Segmentation Mask Representation
                        1. Pixel-level Object Delineation
                          1. Instance Segmentation
                          2. Keypoint-based Representation
                            1. Landmark Tracking
                              1. Skeleton-based Models
                            2. Appearance Models
                              1. Color Histograms
                                1. RGB Color Space
                                  1. HSV Color Space
                                    1. Other Color Spaces
                                      1. Histogram Comparison Metrics
                                      2. Histograms of Oriented Gradients
                                        1. Feature Extraction Process
                                          1. Use in Object Description
                                          2. Deep Features
                                            1. Feature Extraction from Convolutional Layers
                                              1. Transfer Learning for Tracking
                                            2. Motion Models
                                              1. Constant Velocity Model
                                                1. Linear Motion Assumption
                                                  1. Model Parameters
                                                  2. Constant Acceleration Model
                                                    1. Incorporating Acceleration
                                                      1. Use Cases
                                                      2. Kalman Filter
                                                        1. State-space Formulation
                                                          1. Prediction Steps
                                                            1. Correction Steps
                                                              1. Assumptions and Limitations
                                                              2. Particle Filter
                                                                1. Nonlinear and Non-Gaussian Tracking
                                                                  1. Particle Resampling
                                                                    1. Computational Considerations

                                                                Previous

                                                                1. Introduction to Object Tracking

                                                                Go to top

                                                                Next

                                                                3. Single Object Tracking